Although prior studies reveal that custodial grandmothers (CGM) and grandchildren (CGC) face high risk for psychological difficulties, virtually no rigorous studies of psychosocial interventions based on solid conceptual frameworks have been conducted with these vulnerable families. In view of preliminary findings that disrupted parenting mediates the relationship between CGM's psychological distress in the caregiver role and CGC's emotional and behavioral problems, this project involves a randomized clinical trial (RCT) of the comparative feasibility and efficacy of two evidenced-based psychoeducational interventions widely used with other caregiver populations. Grandmothers (N = 504) of CGC (age 5 - 12) will be randomly assigned to one of three conditions: Behavioral Parent Training (BPT; derived from Project KEEP for foster parents; Cognitive Behavioral Therapy (CBT; derived from Coping with Caregiving for caregivers of frail elders); Minimal Support Condition (to control for non-specific treatment factors). The proposed universal interventions include prevention and remediation objectives and involve a group format approach with mental health professionals and grandparent peers serving as co-leaders. The RCT will occur in multiple locations across the US to ensure that findings generalize beyond a single area and that equal numbers of Black (n = 168), Hispanic (n = 168), and White (n = 168) families participate. Prior to the RCT, focus groups will be held separately with CGMs of each race (n = 10 apiece) and experienced practitioners (n = 10) with the aim of assessing the perceived importance and acceptability of recruitment methods, treatment goals, and procedures, including possible differences by race/ethnicity. Multiple-informant and multiple-method assessments during the RCT at pretest, posttest and 6, 12, 18, and 24 month follow-ups will include indicators of CGM psychological distress (anxiety, depression); CGC adjustment (internalizing and externalizing problems); and disrupted parenting (use of ineffective discipline and low nurturance). Based on the conceptual framework of the prominent Family Stress Model, multi-group structural equation modeling analyses will be used to achieve four aims: (a) To compare the short and long-term effectiveness of CBT and BPT; (b) to examine longitudinally the dynamic linkages between CGM parenting practices, CGM' psychological distress, and CGC adjustment as modified by BPT and CBT; (c) to determine if key moderating variables (race/ethnicity, CGC, initial psychological difficulties in CGC and CGM influence RCT outcomes; and (d) to examine key factors related to acceptability and adherence to the proposed interventions. The findings will inform future clinical practice with custodial grandfamilies by identifying which of the proposed interventions are most effective and most acceptable among specific sub-groups (e.g., race/ethnicity; initial risk) and by unraveling the specific antecedent-consequent relations among CGM distress, dysfunctional parenting, and CGC adjustment.